Mysql – Difference between Aurora Global database and Aurora regional database with multiple read replicas

auroraaws-auroramysql-5.6

I am planning to migrate my 300GB RDS PostgreSql database to Aurora MySql using SCT and DMS. RDS Postgresql is in seven regions in current setup. Datapipelines are used to ingest data in these instances and keep them in sync. I was thinking that once I create global database in one region, I will be able to add secondary instances in six other regions. But, I read that global instances support just one additional secondary region.

The only relevant benefits for global database from the docs are:

Having an additional secondary instance will have faster replication
compared to having a read replica.
Faster disaster recovery as the
secondary instance can be promoted to primary under a minute.

Now I am wondering what is the difference between:

Having a Aurora global database in one region with writer and reader, adding a secondary region, adding five read replicas(from primary or secondary)
Having a Aurora regional database (with a writer and reader) and adding six read replicas.

Best Answer

Aurora uses MySQL replication to replicate to a secondary data while Aurora Global database uses underlying hardware to replicate your data from another region which doesn't affect your database performance compared to using read replicas.

Related Solutions

MySQL 5.6 – Difference Between Com_xxx and handler_xxx

Com_delete is number of times a user sent DELETE command to MySQL.

Handler_delete is number of times MySQL server sent delete handler to a storage engine via internal API

To illustrate the difference here's an example:

mysql> delete from t1 where id < 100;
Query OK, 88 rows affected (0.00 sec)

mysql> show status like 'Com_delete';
+---------------+-------+
| Variable_name | Value |
+---------------+-------+
| Com_delete    | 1     |
+---------------+-------+
1 row in set (0.00 sec)

mysql> show status like 'Handler_delete';
+----------------+-------+
| Variable_name  | Value |
+----------------+-------+
| Handler_delete | 88    |
+----------------+-------+
1 row in set (0.00 sec)

MySQL 5.6 Memory Consumption – Troubleshooting High Memory Usage in MySQL Server 5.6

innodb_buffer_pool_size=9G

That accounts for 9GB of usage. That setting does not need to be much bigger than the total amount of data and indexes you have, and should not be much more than about 70% of available RAM.

If the memory usage is bothering you, change that to, say, 100M until you have a lot more data than your current 13M.

(There is a lot of junk at the start of your my.cnf; fortunately, it is overridden by later values.)

Best Answer

Related Solutions

MySQL 5.6 – Difference Between Com_xxx and handler_xxx

MySQL 5.6 Memory Consumption – Troubleshooting High Memory Usage in MySQL Server 5.6

Related Question